The determinants of net interest margins in asean banks in the period 2008 – 2012

Therefore, this study wants to find out the factors that have affected the volatility of NIM after the world economic crisisFigure 5: Trend of Net Interest Margins in Asean banks from 20

VIETNAM – THE NETHERLANDS PROJECT FOR M.A ON DEVELOPMENT ECONOMICS

THE DETERMINANTS OF BANK INTEREST MARGINS IN ASEAN BANKS IN THE PERIOD

2008 – 2012

BY VAN THI THANH NHAN

MASTER OF ARTS IN ECONOMICS OF DEVELOPMENT

HO CHI MINH CITY – DECEMBER 2014

VIETNAM – NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

-THE DETERMINANTS OF NET INTEREST MARGIN IN ASEAN BANKS

IN THE PERIOD 2008 - 2012

By

VAN THI THANH NHAN

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

Ho Chi Minh City, December 2014

VIETNAM – NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

-THE DETERMINANTS OF NET INTEREST MARGIN IN ASEAN BANKS

IN THE PERIOD 2008 - 2012

A thesis submitted in partial fulfilment of the requirements for the degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

VAN THI THANH NHAN

Academic supervisor

Dr NGUYEN TRONG HOAI

Ho Chi Minh City, December 2014

I hereby assure that the thesis “The determinants of net interest margin in Asean banks in the period 2008 – 20012” was made by me under supervisor of Dr Nguyen Trong Hoai

I also certify that data and results of this thesis are honest and have not published

by anyone In addition, the substance of the thesis has not been submitted for any other degrees

VAN THI THANH NHAN

In the process of researching is not always convenient and smooth as expected Sometimes, when I faced with difficulties and challenges I wanted to give up without support, help and encouragement of whose are always by my side Hence, before I introduce the content of this paper I want to send to my loved ones deepest gratitude

To my parents and my big family who are always beside and have instructed me

on my ways whenever I am failure or successful

To my beloved husband who always encourage, motivate me not only in the process of completing this study but also in the life Thank you, love you so much

To my beloved daughter, thanks my angel who has given me the impetus from love and responsibility so that I can awake at night and work hard Thanks for helping me forget all the fatigue and back down when I look at you

Thank to Dr Nguyen Trong Hoai who has always supported me in this process Thanks for your reminder and encouragement so that I could complete this paper Especially, thank for giving me can feel the love and confidential teacher

To the Doctors, tutors in VNP program have imparted knowledge with all of passion, enthusiasm Thanks to the other employees, students in the program support me when I study in VNP program

To the friends, colleagues has supported and facilitated during the completion of this research

VAN THI THANH NHAN

TABLE OF CONTENTS

Certification 3

Acknowledgement 4

List of figures 8

List of tables 9

Abbreviations 10

Abstract 11

CHAPTER 1: INTRODUCTION 1.1 Problem statements 12

1.2 Research objectives 17

1.3 Research questions 18

1.4 Research scope 18

1.5 Research structure 18

CHAPTER 2: LITERATURE REVIEW AND CONCEPTUAL FRAMEWORK 2.1 Literature review for interest margins 19

2.1.1 Definition of net interest margin 19

2.1.2 Determinants of NIM 21

2.1.2.1 Related literature 21

2.1.2.2The macroeconomic factors 23

2.1.2.3The bank specific factors 24

2.1.2.4 The banking market factor 27

2.2 The suggested research approach 32

2.3 The concept framework 34

CHAPTER 3: RESEARCH METHODOLOGY AND DATA COLLECTION 3.1 Identification of variables 36

3.1.1 The dependent variable 36

3.1.2 The independent variables and hypothesis testing 36

3.1.2.1 The macroeconomic factors 36

3.1.2.2 The banking market factor 38

3.1.2.3 The banking specific factors 38

3.2 Data collection and expected results 43

3.3 The research methodology 45

3.3.1 The model 45

3.3.2 The estimation method 45

3.3.2.1 Fixed Effect Model 45

3.3.2.2 Random Effect model 47

3.3.2.3 Selecting the appropriate model 48

3.4 The outline of estimation method 48

CHAPTER 4: DATA ANALYSIS AND DISCUSSION 4 1 Descriptive statistical analysis 50

4.1.1 The data description 50

4.1.2 The summary statistic 51

4.1.3 Testing for correlation relationship 54

4.1.4 Checking for multicollinearity 54

4.1.5 The relationship between independent variables and Net interest margins 55

4 2 Econometric estimation and testing models: 61

4.2.1 Whether FEM or REM is more consistent 62

4.2.2 Fixed Effects Model 63

4.3 Empirical findings 65

4.3.1 Hypothesises rejected 65

4.3.2 Hypothesises accepted 65

CHAPTER 5: CONCLUSION AND RECOMMENDATIONS 5.1 Conclusion 68

5.2 Policy Recommendation 69

5.3 Limitation and further research 71

REFERENCES 72

APPENDIX A 77

APPENDIX B 79

LIST OF FIGURES

FIGURE 1:GDP growth rate in main regions and countries, 2005 – 2009 14

FIGURE 2: The growth rate of worldwide industrial exports, 2005 – 2009 .15

FIGURE 3: GDP growth rate from 2008 to 2009 in Asean countries 16

FIGURE 4: Inflation rate from 2008 to 2012 in Asean countries 16

FIGURE 5: Trend of Net Interest Margins in Asean banks from 2008 – 2012 17

FIGURE 6: The relationship between GDP and NIM 56

FIGURE 7: The relationship between INF and NIM 56

FIGURE 8: The relationship between HHI and NIM 57

FIGURE 9: The relationship between SIZE and NIM 57

FIGURE 10: The relationship between LIQ and NIM 58

FIGURE 11: The relationship between CRD and NIM 58

FIGURE 12: The relationship between CAP and NIM 59

FIGURE 13: The relationship between OPE and NIM 59

FIGURE 14: The relationship between IIP and NIM 60

FIGURE 15: The relationship between MGE and NIM 60

LIST OF TABLES

Table 1: Feature and source of variables 43

Table 2: Data description 50

Table 3: Deterministic statistic of main variables 51

Table 4: Correlation coefficient of variables 54

Table 5: Testing for multicollinearity 55

Table 6: Comparison of regression result of FEM and REM 61

Table 7: Testing for selecting appropriate model 62

Table 8: Results of Fixed Effect Estimator 63

Table A: The summary of main literatures review 28

ASEAN : Association of Southeast Asian Nations 11

GDP : Gross domestic product 17

CEE : Central and Eastern Europe 22

CPI : Consumer Price Index 37

FEM : Fixed Effect Model 11

REM : Random Effect Model 11

WB : Work Bank 11

IMF : International Monetary Fund 14

LSDV : Least Square Dummy Variable 46

FGLS : Feasible Generalized Least – Squares 46

This study investigates the determinants of bank interest margins in Association of Southeast Asian Nations (Asean) banks over the period 2008 – 2012 by using of a panel data set In this research, the Hausman test will be used to make my choice is (FFM) Fixed Effect model or Random Effect model (REM) to analyze impact factors on the net interest margins (NIM) In addition, the data serving for research come from the World Bank (WB) Indicators database and Bankscope database According to the expected results, there will be three main group impacts on bank margins: the bank specific characteristics, the macroeconomic conditions, the banking market characteristics The results do not indicate that macroeconomic factors and structure of market have any impact on net interest margins In the banking characteristics, the paper finds that there is only capital adequacy, implicit interest payment and managerial efficiency effecting on bank margins Based on these findings, the policy makers can advance petitions and policies for managing bank margins effectively following developing strategy of banks

by controlling significant effects on bank margins; namely capital adequacy, implicit interest payment and quality of management

Keyword: Net interest margins, determinants, Asean, bank

CHAPTER 1: INTRODUCTION

1.1 PROBLEM STATEMENTS:

Bank systems play a cruel role in the modern economy, banks are considered as intermediate institutions in social Efficiency of financial intermediaries influences on developing economy with specifics example in India (Sathye, 2003) As a good example, development and efficient of banking system impacted on Hungarian economy in transitional period (Hasan & Marton, 2003) On the other hand, the banking sector plays important role in monetary policy in Europe through deposit and lending channel,

(Kashyap & Stein,1997), this paper proclaim that the degree of bank dependence on customers and ability of bank are main factor on setting up monetary for the Europeanmonetary union The more efficient banking systems, the better economy stand shock Similarly, efficiency of banking system is an important part in stabilizing economy

(Athanasoglou, Brissimis and Delis, 2005) and (Sbracia & Zaghini, 2003) At the same time, the banking sector is seen as a bridge between the different sectors of the economy, the economic bridge between regions and nation together Because of the important role

of the banking system, the performance of the banking system has been considered one of the important issues in the country’s economy in particular and the world economy in general Bank activities in many different fields such as deposits, credit, banking services, money transfers,… and the other finance fields securities, real estate, etc However, the activities which bring main profit for bank are deposit and lending, they are seem a specific areas of banking system Hence, Banks play a crucial role as the financial intermediary

The traditional business operation of bank is that loans and deposits, bank collects funds from customer’s deposit and meet demand loans base on this funds, so when the cost of fund is high, it will effect on bank profitability Net interest margin is supposed a typical indicator for relationship between deposits and loans of bank NIM can be

measured by two ways In the first method, NIM is calculated by difference between loan interest rates and deposit rates According the second way, NIM is measured by the difference between interest income and interest expenses of bank divide by total assets for the period considered, data of interest income and interest expenses, total assets usually get from the financial statement The interest income stands for the result of bank after taxes for the period surveyed, the interest expenses is represented the cost of depositing funds in the current period In general, the second method was used in most previous studies on the NIM such as Ho and Saunders (1981), Angbazo (1997), Saunders and Schumacher (2000), etc and most NIM published by banks is calculated income statement based on the second method

As for the impact of NIM on banking operation, Demirgüç-Kunt and Huizinga (1999) reply that the largely on deposits for fund will have less profitability because covering the cost of funds increase, they also found that stable and efficient bank concentrate on banks margins and profitability, the bank interest margins should not be too large Typical representative of banking cost and efficiency is net interest margins As most of previous studies, the net interest margin is calculated as the gap between interest income and interest expense by percent of total earning assets will influence on bank’s efficiency As a result, NIM can be used an indicator of the banking sector At the same time, an efficient intermediation cost as indicated by a low rates of net as well as reflects the effectiveness of monetary policy and financial stability (Hadad et al, 2003) Bycontrast, high costs would reduce the incentive for economic actors Raharjo et al (2014)proved that the banking sector is heathy and able to create more profit, it will can to withstand shocks and contribute to the stability of financial system In a country, the financial sector is dominated by banking sector, any failure in this sector has an immense implication on the economic growth of the country Hence, as a fact, any bankruptcy or poor performance in banking sector has a contagion effect that can lead to bank runs, crises and bring overall financial crisis and economic tribulations (Ongore and Kusa, 2013) Therefore, this research wants to analyze the determinants of net margin in the

financial period, so that can found significant factor and base on that to improve heath of bank through net interest margins

In the period 2008 -2012, the global economy was under pressure from the financial crisis in the United State at all sectors such as industry, services, finance,…Especially, the banking sector was affected significantly by this crisis For example, bankrupt of several large American banks as Lehman Brothers Bank which was the fourth largest in United State in 2008, the Integra Bank Corp in 2011 and a lots of smaller bank over the period 2008 – 2012 The instability of the largest banking system affected whole of banking system in the world, in which the Asean banks were not exceptional Period

2008 -2009 was considered as the worst economic crisis of the world economy The major economies such as America, Japan and Europe fell into recession, GDP declined dramatically, unemployment rate increased highly, a lot of companies bankrupt,… Some countries fell into negative GDP growth as the EU (-0.5%), Germany (-0.8%), the United States (-0.7%), Japan ( -0.2%)some countries like Russia and only 3 , 5% and China from 10% to 8% in 2008 (Figure 1) In addition, Figure 2 also show that the growth rate of worldwide industrial exports decreased remarkably from 2007 t0 2009

Figure 1: GDP growth rate in main regions and countries, 2005 - 2009

Russia India China

Source: International Monetary Fund (IMF) and Author’s calculation

Figure 2: The growth rate of worldwide industrial exports, 2005 – 2009

Source: Source: IMF and Author’s calculation

Grigor and Salikhov (2009) discussed the main causes of this crisis, of which higheconomic growth rates worldwide since the beginning of the 2000s against the background of serious savings imbalances; Negative real interest rates in developed countries; De facto weakening of financial sector regulation with the sudden expanded use of new financial instruments were main factors of developing the global crises

Fidrmuc and Korhonen (2010) have demonstrated the powerful impact of the global crisis

on business cycles in Asian developing countries It found that there was a negative significant impact between the global crises in 2008 to economic development of OECD countries, namely decline of GDP growth rate and low level of business cycle were results of this crisis in OECD countries Ivashina and Scharfstein (2010), Aisen and Franken (2010) showed the effect of the crisis on a bank credit which was main activities

of banks Period 2010 - 2012, the world economy began to recover from the shock of the crisis still lingers to the crisis

The crisis developed and spread to other Asian countries, including the countries of the ASEAN region In ASEAN countries, Figure 3 showed that GDP growth rate

Worldwide industrial exports

fluctuations during the period 2008 - 2012, including the period 2008 - 2009 GDP slump However, GDP growth rate began to recover over the period 2010 – 2012 Similarly, the inflation rate also declined dramatically during 2008 – 2009 in most Asean countries, and

in 2010 -2012 the inflation rate became more stable (Figure 4)

Figure 3: GDP growth rate from 2008 to 2009 in Asean countries.

Source: Work Bank (WB) and Author’s calculation

Figure 4: Inflation rate from 2008 to 2012 in Asean countries

Source: WB and Author’s calculation

On the other hand, Figure 5 showed the trend of NIM in Asean banks from 2008 to

2012 NIM is seemed as the proxy of banking efficiency Hence, fluctuate of NIM will effect on efficiency, profit of banks In this crisis period, the net interest margins tend to decline in Asean countries The Figure 3 showed the trend of mean NIM in Asean banks

over the period 2008 -2010 The graph shows a downward trend in the epicenter of the crisis period 2008 – 2010 and during the recovery phase of economy from 2010 – 2010, NIM has fluctuated Therefore, this study wants to find out the factors that have affected the volatility of NIM after the world economic crisis

Figure 5: Trend of Net Interest Margins in Asean banks from 2008 – 2012

: Source: Bankscope and Author’s calculation

1.2 RESEARCH OBJECTIVES

The goal of the study:

The main goal of this study is to model and measure the significant determinants

of the net interest margins in Asean banks including ten (10) factors: Gross Domestic Product (GDP) growth rate, Inflation rate, banking market structure (represented by HHI), Bank size, Liquidity risk, Credit risk, Capital adequacy, Operating cost, Implicit interest payments, Managerial efficiency At the same time, to draw empirical conclusion and suggest some policy recommendations for decision makers

Specific objectives:

To meet this goal, specific objectives are set out:

1 Determine the factors, magnitude, sign and significant level of determinants of NIM

2 Inferring conclusions to suggest recommendations

1.3 RESEARCH QUESTIONS

To solve objective of this paper, the relevant questions are answered:

1 What factors influence on the bank interest margins in Asean banks?

2 How those factors impact on the bank interest margins?

3 To recommend general policies for managing bank interest margins of Asean banks Which policy recommendation to manage NIM?

1.4 RESEARCH SCOPE

The study will cover determinants impact on bank interest margins in 9 Asean countries (Brunei, Cambodia, Malaysia, Philippines, Laos, Vietnam, Singapore, Thailand, and Indonesia) from 2008 to 2012 In fact, Asean region have 10 members including Brunei, Cambodia, Malaysia, Philippines, Laos, Vietnam, Singapore, Thailand, Indonesia and Myanmar However, there was a limitation of collecting Myanmar data so it will be eliminated Myanmar from this study

1.5 RESEARCH STRUCTURE

The rest of the research is organized as follows Chapter 1 explain reasons of choose this theme and main goals of research Chapter 2 outlines the literature reviews and conceptual framework on the determinants of net interest margins Chapter 3 describes the research methodology and data Chapter 4 presents the main results of analysis Chapter 5 concludes and gives some policy recommendation

CHAPTER 2: LITERATURE REVIEW AND CONCEPTUAL

FRAMEWORK

2.1 LITERATURE REVIEW FOR INTEREST MARGINS

2.1.1 DEFINITION OF NET INTEREST MARGIN:

NIM is preceded from the relationship between deposits and lending at the bank The bank mobilize funds of depositors by deposit interest rate giving, after that the bank invest these funds by giving loans for borrowers with higher interest rate Net interest margins analysis is one way of measuring the cost of financial intermediation, the difference between the cost of interest paid by the borrower to the bank and depositors received interest income (Brock and Suarez, 2000) So that the bank set up the rates for loans and deposit as follow:

RL= r + b

RD = r – a

In which:

RL : the rate on loans

RD : the rate on deposits

R : risk – free interest rate

a : fees charged on loans

b : fees charged on deposits

And the pure margin is:

R L - R D = a + b

Determinant of net margins can be explained using two approaches, namely the traditional approach and modern approach The traditional approach views of variables that affect the NIM are done by analyzing the balance sheets of bank, whereas the modern approach by taking into account demand and supply rate on the microstructure of

the bank Almost previous studies based on modern approach Therefore, NIM is the ratio of net interest income to total earning assets of banks and this indicator usually published by bank report annually Net interest income is the difference between interest income and interest cost earned on interest expense paid A study conducted by Ho and Saunders (1981) is a pioneer in analyzing the NIM to make banking model as intermediary between recipient and the channeling of funds Ho and Saunders (1981)’smodel argue that the bank faces reinvestment risk at the end of the decision period should the short – term rate fall, hence the fees a and b must compensate the bank for bearing this interest rate Hence the fees a and b are optimal and the optimal interest margin is:

(*)

In which:

s : the difference between lending and deposit rates

/ : bank’s risk neutral spread

Q : size of bank transactions

2 ; the instantaneous variance of the interest rate on deposits and loans

R : the bank management’s coefficient of absolute risk aversion

Ho and Saunder (1986) defined NIM as the spread between the interest revenue on bank assets and interest expenses on bank liabilities as a proportion of average bank assets Similarly, Dietrich, Wanzenried and Cole (2010) showed that the net interest margin is the difference between a bank’s interest income and interest expense expressed

as a percentage of interest earning assets In the study of Raharjo, Hakim, Manurung and Maulana (2014), NIM is measured by the ratio of the net interest income to average total earning assets of banks Net interest income is the difference between interest income and interest costs earned on interest expense paid while the earning assets that are accounted for productive assets that generate interest Besides, net interest margin was specified by

the difference between the cost of interest paid for the borrower and depositors receivedinterest income in Brock and Suarez (2000)’research In general, above definitions have difference phrase but the same meaning about NIM In this study, NIM will be measured

by the ratio of net interest income to total earning assets, while net interest income is calculated by interest income minus interest expenses The database of NIM is provided

by Bankscope so there is an advantage when using Banksope data because of homogeneousness among the database of NIM across countries

2.1.2 DETERMINANTS OF NIM

2.1.2.1 RELATED LITERATURE:

Ho and Saunders (1981) is one of first persons who examine about determinants of net interest margins The result showed that main determinants of NIM includes interest rate volatility, size of transactions, risk aversion and market competition by using the two step regression procedure In the first step, the NIM was estimated by regression on bank specific characteristics and in the second steps, NIM was estimated by macroeconomics and market structures characteristics In the dealership model of Ho and Saunders (1981), the bank plays a role as a risk averse which facing with cost of fund in loans and deposit markets There are many studies based on Ho and Saunders (1981)’s model to analyze net interest margins according to different aspects

Wong (1997) also extended the model of Ho and Saunders (1981) It found that credit and interest rate risk are affected on NIM based on the simple firm – theoretical model This firm – theoretical model was based on behavior of risk-averse bank with NIM under credit risk and interest rate risk The results of this paper gave information about positive relationship between NIM and market power, operating cost ad credit risk; simultaneously, there was a positive impact of interest rate risk on bank interest margins.Relying on the dealership model of Ho and Saunders (1981) , Saunders and Schumacher

(2000) also employed the two step procedure to find impacts on NIM in some countries are that Germany, Italy, Switzerland, UK, Spain, France, US over the period 1988 – 1995 under implicit interest rate, the opportunity cost and credit risk On the other hand, Claeysand Vander Vennet (2008) used random effect estimator to compare different determinant of NIM between Cemtral and Eastern Europe and the West The study examined that interest rate volatility and regulatory restriction such as minimum capital, liquid reserves requirement, implicit interest rate effected on NIM significantly Angbazo (1997) also based on the dealership model of Ho and Sauders (1981), Mcshane and Sharpe (1985) and Allen (1988) to explore the interest rate risk, default risk, liquidity risk and off-balance sheet is consistent with fluctuation of NIM for 1989-1993 with 1400 observations on 286 commercial banks from the Call Report data By contrast, Lin et al.(2012) applied the switching regression model for their research about banks margins

on diversification at banks in some Asian countries as China, India, Indonesia, Japan, Philippines, Singapore, South Korea, Taiwan and Thailand from 1997 to 2005 The findings showed that NIM can be sensitive because of bank risk factors as liquidity risk, interest rate risk, credit risk, implicit interest payment, etc and other factors calculated from balance sheet and income statements As most of other study, Dumičić and Ridzak (2012) also deal with determinants of NIM in Central and Eastern Europe ( CEE) based

on Ho and Saunders (1981) model of period 2000 – 2010 but by fixed effect estimator In the EU, Maudos and Guevara (2004) also extended Ho and Saunder(1981) dealership model to explain factors on the NIM, this paper give result that NIM is consistent with market power, concentration positively but have a negative relationship with interest rate risk, credit risk and operating cost In addition, the study by Kasman, Tunc, Vardar and Okan (2010) included the bank specifics, country – specific market characteristic and macroeconomic conditions to demonstrate NIM is related to those characteristics and consolidation also impact on NIM in new and old EU Hence, the model of Ho and Saunder (1981) is considered as basic model for net interest margins At once, this model also deal with the determinants of NIM relied on Claeys and Vander Vennet(2008) and Lin et al.(2012) as main literatures for analyzing Most previous studies have used panel

data to analyze; some studies used data from banks of a country with over the years such

as Ho and Saunders (1981) employed panel data from American banks over the period

1976 to 1979 Similarly, Entrop, Memmel, Ruprecht and Wilkens, (2012) determine determinants of NIM in Albanian banking system from 2001 to 2007; Fungacova and Poghosyan (2009) consider factors effected on NIM through data from Russian banks inperiod 1999 - 2007.; Williams (2007) also used panel data from 1989 to 2001 to analyze determinants of NIM in Australia As a whole, in this case panel data is time series data However, a lot of papers use data from many different countries at the same few periods

In detail, Saunders and Schumacher (2000) employed data from seven countries to prove the relationship between NIM and implicit interest rate, opportunity cost, credit risk in

1988 – 1995; the same thing, Claeys and Vander Vennet (2008) used panel data of 36 countries of Western and Eastern European from 1994 to 2001 and so on And in this study, the author use panel data in many countries in the period surveyed of 5 years from

2008 to 2012 Many of the previous studies (Claeys & Vander Vennet, 2008; Dumičić & Ridzak, 2012; Kasman et al, 2010) separated factors into different group including macroeconomics factors, banking market specific and bank specifics variables In this paper, independent variables are divided by three group factors including macroeconomics factors, banking market specific and bank specifics variables

2.1.2.2 THE MACROECONOMIC FACTORS:

To consider macroeconomics variables, this group of factors consider factors

which shows how the health of a national economy to influence on NIM There are many indicators, parameters representing the economic situation of a country such as GDP, inflation, exchange rates, interest rate, trade of goods, and so on However, in this study GDP Growth and Inflation rate was chosen as proxies for the macro factors for demonstrating influence of the economy to the NIM Lots of empirical studies have also demonstrated the effects of GDP and Inflation on the NIM; the majority of papers have demonstrated that the GDP impact on NIM (Schwaiger & Liebeg, 2008; Ben Naceur & Goaied, 2008; Ben Naceur & Goaied, 2008; etc.).Meanwhile, regard to inflation,

Dumičić and Ridzak have proved an impact on NIM inflation negatively in CEE Kasman, Tunc, Vardar and Okan (2010)’s study showed that there was a contra – variant impact on NIM Aliaga-Dıaz and Olivero (2005) said that high inflation is associated with higher cost and also higher income The positive relationship posits that bank income increases more with inflation than bank cost This paper uses GDP growth rate and inflation to analyze their impacts on interest rate margins Demirgüç-Kunt and Huizinga (1999) suggested that inflation increases the bank cost, hence, it makes bank interest margins and profitability rise Although inflation have the significant positive coefficient with bank interest margins and profitability but it is low As regard to GDP growth rate, no impact on NIM and profitability (Demirgüç-Kunt & Huizinga, 1999) In addition, the findings of Claeys and Vander Vennet (2008) shows that the high GDP growth rate will have high margins in (CEE) Central and Eastern European as well as inflation also have significant positive impact on margins

2.1.2.3 THE BANK SPECIFIC FACTORS:

With regards to bank-specific variables, this group muster element of banking performance effecting on NIM Normally, the most of variables come from income statement and balance sheet or other reports with parameters However, this paper just employs some variables based on empirical literatures

Ho and Saunders (1981) studied the factors affecting on NIM including bank specifics factors and outside factors such as the macroeconomic characteristics, market structure by the two step regression In the first step, the authors regression relationship between NIM and the elements inside the bank, in a second step the authors analyze the relationship between NIM and other factors such as macro-economic factors, legal and so

on, the results also showed that the banking specifics elements including managerial, risk aversion, the size of the transaction, and the degree of market competition and the variance of interest rate impact on fluctuation of NIM at US banks over the period 1976-1979

In the CEE region, Claeys and Vander (2008) have used the balance sheet and income statement to analyze the immanent factors within the bank as bank capital adequacy, the ratio of equity to assets task, in 36 Countries with 2279 CEE banks totally The author also shows that the more banks operate efficiently, the more the NIM

is lower in countries joined the EU, but not true in countries that do not join EU Das (2013) studied the factors affecting NIM during the global financial crisis of banks in India, the author has classified the factors into three groups, in which there was a bank specifics group including bank size, liquid asset, asset share, non - performing assets, cost inefficiency, capital cushion

Angbazo (1997) detect credit risk and liquidity risk increases NIM at US banks in the research “ commercial ban net interest margins, default risk, interest rate risk and off balance sheet banking” in 1997 From pervious empirical papers together with collecting data, this study have seven independent variables belong to bank – specific characteristics embracing bank size, credit risk, capital adequacy, liquidity risk, operating cost, implicit interest payments , managerial efficiency Some variables found from empirical research but it is difficult to find data for Asean banks so they cannot employ on this paper

- Operating cost

With operating cost, research of Hawtrey and Liang (2008) about the bank interest margins considered that the impact of operating cost is positive statistically significant which implies that banks may impose an extra interest margin to cover the high operating cost Operating cost is by far the most important driver of NIM in the new EU member and candidate countries’ banking system It has a significant positive relationship with net interest margin, which implies that banks may require an extra interest margin to cover the high operating cost (Kasman, Tunc, Vardar & Okan, 2010) Zhou & Wong, 2008) also found that average operating cost remarkably effect on NIM in commercial bank in mainland China

- Implicit interest payments:

About implicit interest payments, Angbazo, L (1997) the increasing of implicit interest payments will make cost increase, so NIM will go up This finding also supported

a lot research about NIM such as Hawtrey, K., & Liang, H (2008); Zhou, K., & Wong,

M C (2008); Maudos, J and de Guevara J.F (2004); Lin, J R., Chung, H., Hsieh, M H.,

& Wu, S (2012) Implicit interest payment is considered a most factors of NIM, because

it directly effects on lending rates and deposit rates of banks Banks give many services for customer but it is not quietly free, cost of these services will be added on rates for customers Hence, there is a positive relationship between IIP and NIM

- Managerial efficiency

According to managerial efficiency measure quality of management, typically, banks which have effectively managed will bring more profit than other banks and soNIM is higher ( Angbazo (1997) Hawtrey, K., & Liang, H (2008)) However, in EU, Kasman, A., Tunc, G., Vardar, G., & Okan, B (2010) found that the bank offer higher deposit rates and lower lending rates when cost for ME increases

2.1.2.4 THE BANKING MARKET FACTOR:

The third factor is banking market variable, measuring the degree of concentration

in the market structure In a lot of papers, the Herfindahl index is used to consider as a proxy for market concentration For instance, Maudos and Guevara (2004)’s study get Herfindahl index as proxy, the result of this research show that NIM is effected positively and highly significant by this index Furthermore, the concentration is one of main determinant impact on interest margin in CEE (Claeys & Vander Vennet, 2008), as other researches, the Herfindahl index represented concentration impacts on margins positively and highly significant Similarly, Demirgüç-Kunt and Huizinga (1999) found that the bank concentration ratio influence on both bank margins and profitability and trend of high margins in large banks

Table A: the summary of main literatures review

No Author Model and Factors considered Methodology Results

Two step regression procedure

Net interest margin is determined by four main factors: risk aversion, size of transactions, degree of market competition and the variance of interest rate.

implicit interest payments;

opportunity cost of non-interest bearing reserves; management efficiency; branching regulation)

Testing hypothesis and F-test computed under null hypothesis

Default risk, interest rate risk were consistent with NIM And management efficiency, opportunity cost effected positive significant ly on NIM while

liquidity risk impacted negatively

of demand and savings deposits in total depotsits; Herfindahl index or comcentration ratio; power market, share of assets/total bank assets;

share of total assets/the assets of the median bank; efficiency ratio, the inverse of total overhead costs/total assets; real GDP growth; the inflatiion rate, changing in CPI; real short term interest rate; the

transition indicator for enterprise reform; the transition indicator for banking reform)

Random – effect panel data estimation approach

Concentration, Operational efficiency, Capital adequacy, Risk behavior and Inflation were main determinants of NIM

(1997).

NIM=(risk aversion; market power;

cost of loans; interbank market rate;

credit risk; interest rate risk; equity capital)

Two step regression procedure

The findings showed that there were the positive relationship between market power, operating cost, credit risk and net margins.

5 Saunders NIM=f(implicit interest rate; Two step Interest rate volatility, regulatory

No Author Model and Factors considered Methodology Results

Herfindahl insex; size; GDP growth;

GDP per capital; state; foreign share;)

Fixed effect model

Credit risk, interest rate risk, operating cost, ownership matter impacted significantly on NIM in CEE from 2000

average size of operation/volume of loans; implicit interest payments;

opportunity costs of bank reserves;

Quality of management )

Single step estimation procedure

There four determinant of NIM were specified including:

Endogenous switching regression model

There are relationships between Management efficiency, capital base, Opportunity cost , Implicit interest payment , Liquidity risk , Interest rate risk and NIM

9 Williams

(2007).

NIM=(market power; operating cost; managerial risk aversion;

interest rate volatility; credit risk;

interaction between interest rate risk and credit risk; credit risk ; implied interest; size of bank operation; the cost of bank reserves; management quality; liquidity risk; control variables)

Fixed effect model and Random effect model

The following factors were considered effect on NIM:

No Author Model and Factors considered Methodology Results

Tunc,

Vardar and

Okan

(2010).

default risk; capital adequacy;

implicit interest payments; size;

managerial efficiency; Lerner index

‘ deposit to equity; GDP growth;

inflation:capitalization)

estimation approach

managerial efficiency, macroeconomics factors effected significant on NIM in

EU members over the period 1995 -2000 and 2001- 2006

Dependent variables: net interest

margins and profitability.

Independent variables:

B: characteristics of bank come from income statement and balance sheet.

X; characteristics of country

Two step estimation with the weighted least squares.

With data from 80 countries over the period 1988-1995, the author indicated that determinants which were bank characteristic, macroeconomics factors, bank taxation, regulation, financial structure, underlying legal and institution indicators have relationship with bank interest margins and bank profitability

E[µ i ] = E[v i,t ] = E[µ v i,i ]

Dependent variable: net interest

margins

Independent variables:

- Bank specific variables: size of bank – logarithm of total earning assets; cost to income ratio; ratio of non-interest revenue `to average assets; total capital ratio; net loans

to total assets; ratio of loans to customer deposits; ratio of impaired loans to gross loans; ratio of reserves for impaired loans to impaired loans; ratio of taxes paid and pre-tax profit; growth rate of gross loans.

- Macroeconomic variables: GDP

Fixed effect estimation.

The results showed that credit risk ,capital adequacy, size, GDP growth rate,

Inflation rate were main determinants of NIM

No Author Model and Factors considered Methodology Results

growth; inflation rate; current account; government debt; GDP per capita; Exchange rate change;

Eurisation dummy; country spread;

3 month money market rate;

volatility of domestic money market.

- Banking markets variables:

NIM=f(Legal origin; creditor right;

contract enforcement days; rule of law; inflation; GDP growth;

operational cost; credit quality;

regulatory capital costs; bank concentration ratio)

Generalized method of moments

The determinants of margins across different countries

Both inflation rate and GDP growth are significantly higher in developing countries when compared with developed countries While the governance variables were more important for developing countries, the bank specific factor are matter for developed countries

scale effects; implicit interest payments; opportunity cost of bank reserves; managerial efficiency).

Random Effect and Fixed Effect

Operating cost is measure by the ratio of operating expenses to average total assets The impact of operating cost is positive and statistically significant which implies that banks may impose an extra interest margin to cover the high operating cost.

2.2 THE SUGGESTED RESEARCH APPROACH:

The study of NIM was researched firstly by Ho and Saunders (1981 In the dealership model of Ho and Saunders (1981), the bank was suggested as a risk-averse

dealer as the financial intermediaries The classical operation of bank is intermediaries between suppliers and demanders of funds through deposit rates and credit rates Banks mobilize fund in the money market from supplier’s funds applied deposit rates and meet demand loans for demanders of funds with loan rates relying on these fund mobilized in the market A risk-averse dealer must face asymmetric arrival of demands for loans and supplies and of deposits (Zhou, K., & Wong, M C (2008).), it means that when a deposit come bank need a loan demand and vice versa On the other hand, period of deposits are not same together and with period of loans, so interest rates set up different and bank must calculate cost of these funds most appreciate Simultaneously, lending and deposit operation should be held in a harmonized manner to banks avoid the risk of interest rates

in the money market Normally, borrowers usually have long-term capital needs while period of deposits are usually shorter so the balance of funds in this operating bank isvery important activity, banks often rely on interest rates to attract capital Besides, banks have to reinvest funds without loans demand to cover the cost of deposits Hence, banks confront with several risks as interest rate risk, default risk, credit risk, Angbazo ( 1997) had extend Ho and Saunders (1981) with adding risk factors on NIM

The dealership model of Ho and Saunders (1981) becomes the basic model for last researchers about net interest margins Angbazo(1997), Saunders and Schumacher (2000), Claeys and Vander Vennet(2008), Maudos and de Guevara (2004) developed this NIM model in their studies Furthermore, Angbazo (1997) and Maudos and de Guevara (2004) extended model of Ho and Saunders (1981) as follow:

? = ? + ? = 1 2 ? ? ??

? + ? ??

?? + 1 2 ? ? (?) ? + ? (? ) ? ?

-14

 ?? (? )? +? (? )? ? : The average operating cost.

 −? "(?? )? (?? ) : The coefficient of absolute risk aversion

 ??2 : the volatility of money market interest rates

 σL2 : the credit risk

 ??? : the covariance between interest rate risk and credit risk

 (? + 2?0): The total volume of credits

 (? + ? ) : The average size of the credit and deposit operations undertaken by the bank

This model analyzes determinants of NIM under risk factors of the spread between lending and deposit rates in the money market Angbazo (1997) explored the more risky loans and higher interest rate, the higher bank interest margins in US commercial banks for 1989 – 1993 NIM is affected by market power, interest rate risk, credit risk and operating cost in EU banks during the period 1993 – 2000

The NIM model of Ho and Saunders (1981) is employed in a lot of empirical research with different aspects Kasman, Tunc, Vardar and Okan, (2010) find the relationship of consolidation and commercial bank net interest margins with evidence from the old and new European Union members and candidate countries Aliaga-Dıaz and Olivero (2005) applied the dealership model of Ho and Saunders to find the cyclical behavior of Net interest margins in US banks On the other hand, the relationship between net interest margin and market interest rates was found in English, W B (2002)’s study In addition, Fungáčová and Poghosyan (2011) proved that NIM concern with ownership factors in Russia Dietrich, Wanzenried and Cole (2010) concluded that net interest margins are different across countries because of bank specific factors and macroeconomics variables

2.3 THE CONCEPT FRAMEWORK:

The concept framework shows the determinant of interest margins include three factor groups as bank specific characteristics; macroeconomics factors; banking market characteristics The first group, macroeconomics condition is considered two main factor GDP growth and inflation rate The next group, bank specific group involves parameters which performance of bank including seven factors are bank size, liquidity risk, credit risk, capital adequacy, operating cost, implicit interest payments, managerial efficiency.With regard to banking market group, the Herfidahl index is deputized to structure market impact on bank margins.

Net Interest Margin

Macroeconomic

characteristics Bank-specific characteristics

SUMMARY CHAPTER 2

The chapter 2 gives an overview on theoretical literature about net interest margins and determinant of net interest margins including GDP growth rate, inflation rate, market structure, scale of bank, liquidity risk, and credit risk, capital adequacy, operating cost, implicit interest payments and managerial efficiency In addition, this chapter specifies the computing method of each variable and presents the appropriate empirical model for this research based on the theoretical models of Ho and Saunders (1981) and Angbazo (1997) Furthermore, the conceptual framework is established on theoretical literature reviews showed

CHAPTER 3: RESEARCH METHODOLOGY AND

DATA COLLECTION

3.1 IDENTIFICATION OF VARIABLES:

3.1.1 THE DEPENDENT VARIABLE:

To approach net interest margin, there are two way In the first method, NIM is measured as the difference between the contractual interest rates for deposits and loans Secondly, NIM is calculated the difference between the interest income and interest expenses in the period considered In case the first method contains more disadvantages then the second, because there are many different sources for deposits and loans at the banks, so they can not correspond with together However, the second also have weak point, Demirgüç-Kunt and Huizinga (1999) reckoned that the interest income and interest expenses have tend to materialize in different periods With the second one, the data will

be based on the financial statements of the banks

In this study, net interest margin has defined as the cost of intermediation measured by the difference the gross paid by the borrower and the interest income received by depositors (Bernake, 1983 and Brock & Suarez, 2000) As a lot of empirical papers, NIM is calculated as the difference between interest income and interest expenses

as a proportion of total earning assets (Claeys & Vander Vennet, 2008; Dietrich, A., Wanzenried & Cole, 2010; Entrop, Memmel, Ruprecht, & Wilkens, 2012; ect ) Data for NIM will be gotten from the Bankscope

3.1.2 THE INDEPENDENT VARIABLES AND HYPOTHESIS TESTING:

3.1.2.1 THE MACROECONOMIC FACTORS:

 GDP growth rate (GDP) variable belongs to macroeconomic factor; GDP is

measured by the GDP per capital growth rate GDP growth rate represents the

economic growth of each country Changing of economic growth will effected on prices, cost and business cycle because of financial monetary policies changed to control the market Therefore, economic growth may affect in NIM Most previous studies have same results is that there the positive relationship between GDP and NIM In Western European bank markets, Claeys and Vander Vennet (2008) showed that there is the positive association between business cycle and NIM The results reflect the higher economic growth is related with higher net interest margin The higher GDP growth give more opportunities and more demand for credit and make increase the bank margins (Dumičić & Ridzak, 2012) In contrast, Ben Naceur and Goaied (2008) has proved that there is not relationship between economic growth and NIM in Tunisia similar to Ben – Kediri et al (2005)‘s finding Data for GDP growth rate from 2008 to 2012 of Asean countries in this research come from World Bank

Hypothesis 1: Economic growth (GDP) is expected a positive significant impact

on NIM The greater economic growth will have a higher net interest margins

 Inflation rate (INF) is inflation rate calculated by the changing of Consumer

Price Index (CPI) Inflation rate is considered as a factor effecting prices of markets Purchasing power in the market will decrease when inflation rise, normally, high inflation concern high interest rate of deposits and loans Hence, inflation is suggested as macroeconomic element effect on changing of NIM In Kasman et al (2010), the increase of inflation rate make cost and income be higher, there was a positive relationship between inflation rate and NIM In Indonesia, cover the period 2008 – 2012 too, Raharjo, Hakim, Manurung and Maulana (2014) found that inflation has a positive significant effect on NIM

Hence, inflation rate can be positive and negative sign in this model As GDP growth rate, inflation data will get from World Bank

Hypothesis 2: Inflation rate is expected that there is a positive effect on bank interest margin.

3.1.2.2 THE BANKING MARKET FACTOR:

 Herfindahl - Hirshman index (HHI): Market structure in the banking sector is a

factor measured by its size distribution index, the position of each bank in the whole banking market in each period For this study, HHI is considered as proxy

of the market structure HHI is defined square of bank asset share in loan market The number of previous study demonstrate that the Herfindahl index has a highly significant and positive link with NIM such as Claeys &Vander Vennet, 2008; Fungacova & Poghosyan, 2009 and Maudos & Guevara, 2004 Therefore, the coefficient sign of Herfindahl index can be positive with NIM

Hypothesis 3: HHI is expected that there is a positive impact on net interest margins

3.1.2.3 THE BANKING SPECIFIC FACTORS:

 Bank size (SIZE) measure the size of bank based on the logarithm of total assets

Bank size delegate an operating scale of banks in market Normally, the banks which have larger size will have lower bank margin and vice versa the small bank apply higher margins because of the higher interest rate for borrowers Finding of Fungacova and Poghosyan (2009) also support above argument that operating size

of bank effect on NIM negatively Similarly, Dumičić and Ridzak (2012) showedthat in CEE banks the bigger banks will have the lower cost of income and NIM will be higher As a result, the coefficient of SIZE variable is expected positive sign with NIM

Hypothesis 4: It is expected that scale of bank will effect on net margins significantly by a negative relationship

 Liquidity risk (LIQ) is liquidity risk measured by the ratio of liquid assets to total

liability Liquidity risk is the situation bank do not meet demand enough cash for withdrawing cash of depositor or new loans, no any bank want high liquidity risk, the liability always is ensured by the liquid assets The coefficient of liquidity risk

is expected significantly negative with NIM The more demand liabilities of the bank are backed up by liquid assets, the lower the liquidity risk of the bank and its margins in Russian Banks from 1999 to 2007 (Fungacova & Poghosyan, 2009) In

1997, Angbazo developed the dealership model of Ho and Saunder (1981) adding risky factors, including liquidity risks; he discovered that the liquid assets increase the liquidity risk in bank margins decreases With the research “ the cyclical behavior of net interest margins: Evidence from the United State Banking Sectore”, Aliaga-Dıaz and Olivero (2005) find evidence that the counter cyclicality of balance sheet liquidity is explained by that, credit risk increase more for risky and illiquid assets than for more liquid assets It is result of previousstudies; the coefficient of liquidity risk is expected significantly negative with NIM in this research

Hypothesis 5: Author expects that liquidity risk will have a negative effect on net interest margin.

 Credit risk (CRD) As we analyzed, the credit risk is the most important

determinant of NIM and it is employed by the ratio between total loans and total assets The sign of credit risk is positive as a lot of researchers demonstrated Wong (1997) explored the determinants of bank margins under credit risk; the bank interest margins have positive relationship with credit risk It means that when percentage of total loans to totals assets goes up, the interest spread tend to